Accent: the Elephant in the Room

Understanding spoken English is challenging. Learners often complain that the language they hear out in the real world is completely different from the English they heard in the classroom. All the rules of correct grammar and pronunciation seem to go out of the window. In the ELT community, this problem has been recognised. Nowadays, teachers often focus on those aspects of pronunciation which are known as connected speech – things like cutting some of the sounds, or blending them together. However, there is another aspect of spoken English which is relatively ignored, and that is accent. It is often assumed that learners should follow either a standard American or British model in their own spoken production, but little attention is given to the myriad accents they might encounter as listeners. Perhaps this topic is just too difficult to deal with, so we say little or nothing about it. It is ‘the elephant in the room’.

Let’s consider what it is about accent that makes it so difficult to deal with. I think there are three main issues: Firstly, the sheer variety of accents in English as a global language. Secondly, the complex technical terminology used in texts on the subject. And thirdly, the practical difficulty of finding material to deal with this in the classroom. I will focus on each of these issues in turn.

1 Sheer variety of accents of English

We sometimes talk as if there were only two accents – American or British, but in fact, as any visitor will have noticed, there are many accents in those countries. And it seems that the closer we look, the more we find. For example, we can identify a Northern English accent, but within that, we would find more, such as a Liverpool accent or a Hull accent. Then there are accents from other English-speaking countries such as Australia or Jamaica. Then there are accents of English from non-English-speaking countries such as China or France. Plus, there are blended forms, such as multicultural London accents. If we attempt to deal with all of the above in turn, we will have enough on the syllabus to occupy a very long course indeed! 

The solution, I think, is this: instead of teaching learners all of the accents, we focus on individual pronunciation features which cut across all accents around the world. For example, one such feature would be the letter R (where there is no vowel after it) and whether or not it is pronounced. This feature-based approach could be supplemented by focussing on those few specific accents which are most likely to be encountered by our learners, on a need-to-know basis.

2 The complexity of terminology in describing accents

It is easy enough to hear that one accent is different from another. What is not so easy is to express that difference in words. This problem is particularly acute in the case of vowel sounds: how can we talk about these without specialist jargon and the use of the full phonetic alphabet (not just the phonemic charts which are familiar to many teachers)? I think it is possible to do this, provided we accept that our descriptions will not be as precise and scientific as those in academic texts. Instead of using phonetic symbols, we can use keywords (words representing vowel sounds in different accents), as has been effectively done by John Wells (Wells, 1982). We can also compare accents with one another using homophones. These may be words which sound identical in one accent but not in another, such as court and caught. Alternatively, they may be cross-language homophones – for example, mate in Accent A sounds exactly like might in Accent B.

3 Finding classroom materials

It is easy enough to find audio material in a range of different accents on the internet. What is not so easy is finding the same text in different accents. This is important, because only if we have the same text can we compare like with like. This is necessary if we want to focus on specific pronunciation features, as I suggested above. Fortunately, we can now turn to technology to help in this matter. We can paste short texts into online text-to-speech tools such as Elevenlabs, Natural Reader or Luvvoice, and have them read aloud by voices in a huge variety of accents, all generated by A.I. The quality is not always convincing in terms of intonation, but apart from that, specific pronunciation features found across different accents can be very clearly demonstrated. This is the strategy I have used in my book Accent for Listeners (Hancock, 2026).

In this article, I have explained why accent is important in language teaching. I have argued that like an elephant in the room, it is something which looms large, and yet nobody wants to talk about it because it is so difficult. Finally, I have suggested how the difficulty can be overcome by changing our strategy and by using technological support. Perhaps now we can stop ignoring the elephant!

About the author

Mark Hancock is an ELT author and trainer. His books include Pronunciation Games, English Pronunciation in Use, and the PronPack series. He lectures at the University of Chester and online. For more on the PronPack series, go to https://pronpack.com/

Hancock, M. (2026), Accent for Listeners Hancock McDonald ELT

Wells, J.C. (1982), Accents of English CUP

Elevenlabs, https://elevenlabs.io/app/speech-synthesis/text-to-speech

Natural Reader, https://www.naturalreaders.com/

Luvvoice, https://luvvoice.com/